Mercurial > octave
changeset 21451:763e30285935
eliminate configuration macros from oct-fftw.h (bug #41027)
* oct-fftw.h: Eliminate FFTW configuration macros.
Don't include fftw.h.
* oct-fftw.h, oct-fftw.cc (octave_fftw_planner,
octave_float_fftw_planner): Use void pointer to represent fftw_plan
and fftwf_plan and cast as needed.
* fftw.cc: Include fftw.h
author | John W. Eaton <jwe@octave.org> |
---|---|
date | Tue, 15 Mar 2016 14:58:19 -0400 |
parents | 2868abbc88eb |
children | 769f9a7c02ae |
files | libinterp/dldfcn/fftw.cc liboctave/numeric/oct-fftw.cc liboctave/numeric/oct-fftw.h |
diffstat | 3 files changed, 133 insertions(+), 119 deletions(-) [+] |
line wrap: on
line diff
--- a/libinterp/dldfcn/fftw.cc Mon Mar 14 23:22:32 2016 -0400 +++ b/libinterp/dldfcn/fftw.cc Tue Mar 15 14:58:19 2016 -0400 @@ -26,6 +26,10 @@ #include <algorithm> +#if defined (HAVE_FFTW3_H) +# include <fftw3.h> +#endif + #include "oct-fftw.h" #include "defun-dld.h"
--- a/liboctave/numeric/oct-fftw.cc Mon Mar 14 23:22:32 2016 -0400 +++ b/liboctave/numeric/oct-fftw.cc Tue Mar 15 14:58:19 2016 -0400 @@ -29,6 +29,10 @@ #include <iostream> #include <vector> +#if defined (HAVE_FFTW3_H) +# include <fftw3.h> +#endif + #include "lo-error.h" #include "oct-fftw.h" #include "quit.h" @@ -61,7 +65,7 @@ octave_fftw_planner::octave_fftw_planner (void) : meth (ESTIMATE), rplan (0), rd (0), rs (0), rr (0), rh (0), rn (), - rsimd_align (false) + rsimd_align (false), nthreads (1) { plan[0] = plan[1] = 0; d[0] = d[1] = s[0] = s[1] = r[0] = r[1] = h[0] = h[1] = 0; @@ -74,8 +78,8 @@ if (! init_ret) (*current_liboctave_error_handler) ("Error initializing FFTW threads"); - //Use number of processors available to the current process - //This can be later changed with fftw ("threads", nthreads) + // Use number of processors available to the current process + // This can be later changed with fftw ("threads", nthreads). nthreads = num_processors (NPROC_CURRENT); fftw_plan_with_nthreads (nthreads); #endif @@ -88,15 +92,15 @@ { fftw_plan *plan_p; - plan_p = &rplan; + plan_p = reinterpret_cast<fftw_plan *> (&rplan); if (*plan_p) fftw_destroy_plan (*plan_p); - plan_p = &plan[0]; + plan_p = reinterpret_cast<fftw_plan *> (&plan[0]); if (*plan_p) fftw_destroy_plan (*plan_p); - plan_p = &plan[1]; + plan_p = reinterpret_cast<fftw_plan *> (&plan[1]); if (*plan_p) fftw_destroy_plan (*plan_p); } @@ -121,10 +125,27 @@ return retval; } -#define CHECK_SIMD_ALIGNMENT(x) \ +void +octave_fftw_planner::threads (int nt) +{ +#if defined (HAVE_FFTW3F_THREADS) + if (instance_ok () && nt != threads ()) + { + instance->nthreads = nt; + fftw_plan_with_nthreads (nt); + // Clear the current plans. + instance->rplan = instance->plan[0] = instance->plan[1] = 0; + } +#else + (*current_liboctave_warning_handler) + ("unable to change number of threads without FFTW thread support"); +#endif +} + +#define CHECK_SIMD_ALIGNMENT(x) \ (((reinterpret_cast<ptrdiff_t> (x)) & 0xF) == 0) -fftw_plan +void * octave_fftw_planner::do_create_plan (int dir, const int rank, const dim_vector dims, octave_idx_type howmany, @@ -133,7 +154,7 @@ const Complex *in, Complex *out) { int which = (dir == FFTW_FORWARD) ? 0 : 1; - fftw_plan *cur_plan_p = &plan[which]; + fftw_plan *cur_plan_p = reinterpret_cast<fftw_plan *> (&plan[which]); bool create_new_plan = false; bool ioalign = CHECK_SIMD_ALIGNMENT (in) && CHECK_SIMD_ALIGNMENT (out); bool ioinplace = (in == out); @@ -249,14 +270,14 @@ return *cur_plan_p; } -fftw_plan +void * octave_fftw_planner::do_create_plan (const int rank, const dim_vector dims, octave_idx_type howmany, octave_idx_type stride, octave_idx_type dist, const double *in, Complex *out) { - fftw_plan *cur_plan_p = &rplan; + fftw_plan *cur_plan_p = reinterpret_cast<fftw_plan *> (&rplan); bool create_new_plan = false; bool ioalign = CHECK_SIMD_ALIGNMENT (in) && CHECK_SIMD_ALIGNMENT (out); @@ -385,11 +406,11 @@ { meth = _meth; if (rplan) - fftw_destroy_plan (rplan); + fftw_destroy_plan (reinterpret_cast<fftw_plan> (rplan)); if (plan[0]) - fftw_destroy_plan (plan[0]); + fftw_destroy_plan (reinterpret_cast<fftw_plan> (plan[0])); if (plan[1]) - fftw_destroy_plan (plan[1]); + fftw_destroy_plan (reinterpret_cast<fftw_plan> (plan[1])); rplan = plan[0] = plan[1] = 0; } } @@ -402,7 +423,7 @@ octave_float_fftw_planner::octave_float_fftw_planner (void) : meth (ESTIMATE), rplan (0), rd (0), rs (0), rr (0), rh (0), rn (), - rsimd_align (false) + rsimd_align (false), nthreads (1) { plan[0] = plan[1] = 0; d[0] = d[1] = s[0] = s[1] = r[0] = r[1] = h[0] = h[1] = 0; @@ -415,8 +436,8 @@ if (! init_ret) (*current_liboctave_error_handler) ("Error initializing FFTW3F threads"); - //Use number of processors available to the current process - //This can be later changed with fftw ("threads", nthreads) + // Use number of processors available to the current process + // This can be later changed with fftw ("threads", nthreads). nthreads = num_processors (NPROC_CURRENT); fftwf_plan_with_nthreads (nthreads); #endif @@ -429,15 +450,15 @@ { fftwf_plan *plan_p; - plan_p = &rplan; + plan_p = reinterpret_cast<fftwf_plan *> (&rplan); if (*plan_p) fftwf_destroy_plan (*plan_p); - plan_p = &plan[0]; + plan_p = reinterpret_cast<fftwf_plan *> (&plan[0]); if (*plan_p) fftwf_destroy_plan (*plan_p); - plan_p = &plan[1]; + plan_p = reinterpret_cast<fftwf_plan *> (&plan[1]); if (*plan_p) fftwf_destroy_plan (*plan_p); } @@ -462,7 +483,24 @@ return retval; } -fftwf_plan +void +octave_float_fftw_planner::threads (int nt) +{ +#if defined (HAVE_FFTW3F_THREADS) + if (instance_ok () && nt != threads ()) + { + instance->nthreads = nt; + fftwf_plan_with_nthreads (nt); + // Clear the current plans. + instance->rplan = instance->plan[0] = instance->plan[1] = 0; + } +#else + (*current_liboctave_warning_handler) + ("unable to change number of threads without FFTW thread support"); +#endif +} + +void * octave_float_fftw_planner::do_create_plan (int dir, const int rank, const dim_vector dims, octave_idx_type howmany, @@ -472,7 +510,7 @@ FloatComplex *out) { int which = (dir == FFTW_FORWARD) ? 0 : 1; - fftwf_plan *cur_plan_p = &plan[which]; + fftwf_plan *cur_plan_p = reinterpret_cast<fftwf_plan *> (&plan[which]); bool create_new_plan = false; bool ioalign = CHECK_SIMD_ALIGNMENT (in) && CHECK_SIMD_ALIGNMENT (out); bool ioinplace = (in == out); @@ -588,7 +626,7 @@ return *cur_plan_p; } -fftwf_plan +void * octave_float_fftw_planner::do_create_plan (const int rank, const dim_vector dims, octave_idx_type howmany, @@ -596,7 +634,7 @@ octave_idx_type dist, const float *in, FloatComplex *out) { - fftwf_plan *cur_plan_p = &rplan; + fftwf_plan *cur_plan_p = reinterpret_cast<fftwf_plan *> (&rplan); bool create_new_plan = false; bool ioalign = CHECK_SIMD_ALIGNMENT (in) && CHECK_SIMD_ALIGNMENT (out); @@ -725,11 +763,11 @@ { meth = _meth; if (rplan) - fftwf_destroy_plan (rplan); + fftwf_destroy_plan (reinterpret_cast<fftwf_plan> (rplan)); if (plan[0]) - fftwf_destroy_plan (plan[0]); + fftwf_destroy_plan (reinterpret_cast<fftwf_plan> (plan[0])); if (plan[1]) - fftwf_destroy_plan (plan[1]); + fftwf_destroy_plan (reinterpret_cast<fftwf_plan> (plan[1])); rplan = plan[0] = plan[1] = 0; } } @@ -824,8 +862,9 @@ dist = (dist < 0 ? npts : dist); dim_vector dv (npts, 1); - fftw_plan plan = octave_fftw_planner::create_plan (1, dv, nsamples, - stride, dist, in, out); + void *vplan = octave_fftw_planner::create_plan (1, dv, nsamples, + stride, dist, in, out); + fftw_plan plan = reinterpret_cast<fftw_plan> (vplan); fftw_execute_dft_r2c (plan, (const_cast<double *>(in)), reinterpret_cast<fftw_complex *> (out)); @@ -844,9 +883,10 @@ dist = (dist < 0 ? npts : dist); dim_vector dv (npts, 1); - fftw_plan plan = octave_fftw_planner::create_plan (FFTW_FORWARD, 1, dv, + void *vplan = octave_fftw_planner::create_plan (FFTW_FORWARD, 1, dv, nsamples, stride, dist, in, out); + fftw_plan plan = reinterpret_cast<fftw_plan> (vplan); fftw_execute_dft (plan, reinterpret_cast<fftw_complex *> (const_cast<Complex *>(in)), @@ -863,9 +903,10 @@ dist = (dist < 0 ? npts : dist); dim_vector dv (npts, 1); - fftw_plan plan = octave_fftw_planner::create_plan (FFTW_BACKWARD, 1, dv, + void *vplan = octave_fftw_planner::create_plan (FFTW_BACKWARD, 1, dv, nsamples, stride, dist, in, out); + fftw_plan plan = reinterpret_cast<fftw_plan> (vplan); fftw_execute_dft (plan, reinterpret_cast<fftw_complex *> (const_cast<Complex *>(in)), @@ -892,8 +933,9 @@ octave_idx_type offset = (dv.numel () / dv(0)) * ((dv(0) - 1) / 2); - fftw_plan plan = octave_fftw_planner::create_plan (rank, dv, 1, 1, dist, - in, out + offset); + void *vplan = octave_fftw_planner::create_plan (rank, dv, 1, 1, dist, + in, out + offset); + fftw_plan plan = reinterpret_cast<fftw_plan> (vplan); fftw_execute_dft_r2c (plan, (const_cast<double *>(in)), reinterpret_cast<fftw_complex *> (out+ offset)); @@ -913,8 +955,9 @@ for (int i = 0; i < rank; i++) dist *= dv(i); - fftw_plan plan = octave_fftw_planner::create_plan (FFTW_FORWARD, rank, - dv, 1, 1, dist, in, out); + void *vplan = octave_fftw_planner::create_plan (FFTW_FORWARD, rank, + dv, 1, 1, dist, in, out); + fftw_plan plan = reinterpret_cast<fftw_plan> (vplan); fftw_execute_dft (plan, reinterpret_cast<fftw_complex *> (const_cast<Complex *>(in)), @@ -931,8 +974,9 @@ for (int i = 0; i < rank; i++) dist *= dv(i); - fftw_plan plan = octave_fftw_planner::create_plan (FFTW_BACKWARD, rank, - dv, 1, 1, dist, in, out); + void *vplan = octave_fftw_planner::create_plan (FFTW_BACKWARD, rank, + dv, 1, 1, dist, in, out); + fftw_plan plan = reinterpret_cast<fftw_plan> (vplan); fftw_execute_dft (plan, reinterpret_cast<fftw_complex *> (const_cast<Complex *>(in)), @@ -953,9 +997,10 @@ dist = (dist < 0 ? npts : dist); dim_vector dv (npts, 1); - fftwf_plan plan = octave_float_fftw_planner::create_plan (1, dv, nsamples, - stride, dist, - in, out); + void *vplan = octave_float_fftw_planner::create_plan (1, dv, nsamples, + stride, dist, + in, out); + fftwf_plan plan = reinterpret_cast<fftwf_plan> (vplan); fftwf_execute_dft_r2c (plan, (const_cast<float *>(in)), reinterpret_cast<fftwf_complex *> (out)); @@ -974,10 +1019,11 @@ dist = (dist < 0 ? npts : dist); dim_vector dv (npts, 1); - fftwf_plan plan = octave_float_fftw_planner::create_plan (FFTW_FORWARD, 1, - dv, nsamples, - stride, dist, - in, out); + void *vplan = octave_float_fftw_planner::create_plan (FFTW_FORWARD, 1, + dv, nsamples, + stride, dist, + in, out); + fftwf_plan plan = reinterpret_cast<fftwf_plan> (vplan); fftwf_execute_dft (plan, reinterpret_cast<fftwf_complex *> (const_cast<FloatComplex *>(in)), @@ -994,10 +1040,11 @@ dist = (dist < 0 ? npts : dist); dim_vector dv (npts, 1); - fftwf_plan plan = octave_float_fftw_planner::create_plan (FFTW_BACKWARD, 1, - dv, nsamples, - stride, dist, - in, out); + void *vplan = octave_float_fftw_planner::create_plan (FFTW_BACKWARD, 1, + dv, nsamples, + stride, dist, + in, out); + fftwf_plan plan = reinterpret_cast<fftwf_plan> (vplan); fftwf_execute_dft (plan, reinterpret_cast<fftwf_complex *> (const_cast<FloatComplex *>(in)), @@ -1024,9 +1071,10 @@ octave_idx_type offset = (dv.numel () / dv(0)) * ((dv(0) - 1) / 2); - fftwf_plan plan = octave_float_fftw_planner::create_plan (rank, dv, 1, 1, - dist, in, - out + offset); + void *vplan = octave_float_fftw_planner::create_plan (rank, dv, 1, 1, + dist, in, + out + offset); + fftwf_plan plan = reinterpret_cast<fftwf_plan> (vplan); fftwf_execute_dft_r2c (plan, (const_cast<float *>(in)), reinterpret_cast<fftwf_complex *> (out+ offset)); @@ -1046,9 +1094,10 @@ for (int i = 0; i < rank; i++) dist *= dv(i); - fftwf_plan plan = octave_float_fftw_planner::create_plan (FFTW_FORWARD, - rank, dv, 1, 1, - dist, in, out); + void *vplan = octave_float_fftw_planner::create_plan (FFTW_FORWARD, + rank, dv, 1, 1, + dist, in, out); + fftwf_plan plan = reinterpret_cast<fftwf_plan> (vplan); fftwf_execute_dft (plan, reinterpret_cast<fftwf_complex *> (const_cast<FloatComplex *>(in)), @@ -1065,9 +1114,10 @@ for (int i = 0; i < rank; i++) dist *= dv(i); - fftwf_plan plan = octave_float_fftw_planner::create_plan (FFTW_BACKWARD, + void *vplan = octave_float_fftw_planner::create_plan (FFTW_BACKWARD, rank, dv, 1, 1, dist, in, out); + fftwf_plan plan = reinterpret_cast<fftwf_plan> (vplan); fftwf_execute_dft (plan, reinterpret_cast<fftwf_complex *> (const_cast<FloatComplex *>(in)),
--- a/liboctave/numeric/oct-fftw.h Mon Mar 14 23:22:32 2016 -0400 +++ b/liboctave/numeric/oct-fftw.h Tue Mar 15 14:58:19 2016 -0400 @@ -27,15 +27,9 @@ #include <cstddef> -#if defined (HAVE_FFTW3_H) -# include <fftw3.h> -#endif - #include "oct-cmplx.h" #include "dim-vector.h" -#if defined (HAVE_FFTW) - class OCTAVE_API octave_fftw_planner @@ -60,31 +54,27 @@ static bool instance_ok (void); - static fftw_plan + static void * create_plan (int dir, const int rank, const dim_vector dims, octave_idx_type howmany, octave_idx_type stride, octave_idx_type dist, const Complex *in, Complex *out) { - static fftw_plan dummy; - return instance_ok () ? instance->do_create_plan (dir, rank, dims, howmany, stride, dist, in, out) - : dummy; + : 0; } - static fftw_plan + static void * create_plan (const int rank, const dim_vector dims, octave_idx_type howmany, octave_idx_type stride, octave_idx_type dist, const double *in, Complex *out) { - static fftw_plan dummy; - return instance_ok () ? instance->do_create_plan (rank, dims, howmany, stride, dist, in, out) - : dummy; + : 0; } static FftwMethod method (void) @@ -101,23 +91,12 @@ return instance_ok () ? instance->do_method (_meth) : dummy; } -#if defined (HAVE_FFTW3F_THREADS) - static void threads (int _nthreads) - { - if (instance_ok () && _nthreads != threads ()) - { - instance->nthreads = _nthreads; - fftw_plan_with_nthreads (_nthreads); - //Clear the current plans - instance->rplan = instance->plan[0] = instance->plan[1] = 0; - } - } + static void threads (int nt); - static int threads () + static int threads (void) { return instance_ok () ? instance->nthreads : 0; } -#endif private: @@ -131,13 +110,13 @@ static void cleanup_instance (void) { delete instance; instance = 0; } - fftw_plan + void * do_create_plan (int dir, const int rank, const dim_vector dims, octave_idx_type howmany, octave_idx_type stride, octave_idx_type dist, const Complex *in, Complex *out); - fftw_plan + void * do_create_plan (const int rank, const dim_vector dims, octave_idx_type howmany, octave_idx_type stride, octave_idx_type dist, const double *in, Complex *out); @@ -151,7 +130,7 @@ // FIXME: perhaps this should be split into two classes? // Plan for fft and ifft of complex values - fftw_plan plan[2]; + void *plan[2]; // dist octave_idx_type d[2]; @@ -172,7 +151,7 @@ bool inplace[2]; // Plan for fft of real values - fftw_plan rplan; + void *rplan; // dist octave_idx_type rd; @@ -191,10 +170,9 @@ bool rsimd_align; -#if defined (HAVE_FFTW3_THREADS) - //number of threads when compiled with Multi-threading support + // number of threads. Always 1 unless compiled with Multi-threading + // support. int nthreads; -#endif }; class @@ -221,31 +199,27 @@ static bool instance_ok (void); - static fftwf_plan + static void * create_plan (int dir, const int rank, const dim_vector dims, octave_idx_type howmany, octave_idx_type stride, octave_idx_type dist, const FloatComplex *in, FloatComplex *out) { - static fftwf_plan dummy; - return instance_ok () ? instance->do_create_plan (dir, rank, dims, howmany, stride, dist, in, out) - : dummy; + : 0; } - static fftwf_plan + static void * create_plan (const int rank, const dim_vector dims, octave_idx_type howmany, octave_idx_type stride, octave_idx_type dist, const float *in, FloatComplex *out) { - static fftwf_plan dummy; - return instance_ok () ? instance->do_create_plan (rank, dims, howmany, stride, dist, in, out) - : dummy; + : 0; } static FftwMethod method (void) @@ -262,23 +236,12 @@ return instance_ok () ? instance->do_method (_meth) : dummy; } -#if defined (HAVE_FFTW3F_THREADS) - static void threads (int _nthreads) - { - if (instance_ok () && _nthreads != threads ()) - { - instance->nthreads = _nthreads; - fftwf_plan_with_nthreads (_nthreads); - //Clear the current plans - instance->rplan = instance->plan[0] = instance->plan[1] = 0; - } - } + static void threads (int nt); - static int threads () + static int threads (void) { return instance_ok () ? instance->nthreads : 0; } -#endif private: @@ -292,13 +255,13 @@ static void cleanup_instance (void) { delete instance; instance = 0; } - fftwf_plan + void * do_create_plan (int dir, const int rank, const dim_vector dims, octave_idx_type howmany, octave_idx_type stride, octave_idx_type dist, const FloatComplex *in, FloatComplex *out); - fftwf_plan + void * do_create_plan (const int rank, const dim_vector dims, octave_idx_type howmany, octave_idx_type stride, octave_idx_type dist, const float *in, FloatComplex *out); @@ -312,7 +275,7 @@ // FIXME: perhaps this should be split into two classes? // Plan for fft and ifft of complex values - fftwf_plan plan[2]; + void *plan[2]; // dist octave_idx_type d[2]; @@ -333,7 +296,7 @@ bool inplace[2]; // Plan for fft of real values - fftwf_plan rplan; + void *rplan; // dist octave_idx_type rd; @@ -352,10 +315,9 @@ bool rsimd_align; -#if defined (HAVE_FFTW3F_THREADS) - //number of threads when compiled with Multi-threading support + // number of threads. Always 1 unless compiled with Multi-threading + // support. int nthreads; -#endif }; class @@ -403,5 +365,3 @@ }; #endif - -#endif